Download The P25 component of Galleria silk

Mol Gen Genet (1998) 257: 264±270
Ó Springer-Verlag 1998
ORIGINAL PAPER
M. ZÏurovec á D. KodrõÂ k á C. Yang á F. Sehnal
K. Scheller
The P25 component of Galleria silk
Received: 4 June 1997 / Accepted: 5 September 1997
Abstract The water-insoluble core of lepidopteran silk
is composed of four major proteins, but only three genes
have been identi®ed. This study demonstrates that the
29- and 30-kDa components of Galleria mellonella silk
are derived from a single gene designated P25. The gene
is expressed exclusively in the posterior section of the
silk glands as a 2-kb mRNA, which accumulates in the
feeding larvae and declines at molting. The mRNA encodes a peptide of 24 864 Da that exhibits 51% identity
with the putative product of the P25 gene of Bombyx.
The conservation of several amino acid stretches, including the relative positions of all 8 cysteines in the
mature polypeptide, implies that the P25 proteins play
similar, and apparently signi®cant roles in silk formation
in the two species. A Galleria P25 cDNA yields a peptide
of about 25 kDa when translated in vitro; the 29- and
30-kDa forms present in the silk are derived from this
primary translation product by di€erential glycosylation.
Key words Fibroin á Glycosylation á DNA
sequence á Evolution á Silkworm
Introduction
Several groups of arthropods possess the ability to secrete insoluble ®bres known as silk. Larvae of many
Lepidoptera produce several types of proteins that are
assembled into the ®rm core and sticky coating of the
Communicated by K. Illmensee
M. ZÏurovec á D. KodrõÂ k á C. Yang á F. Sehnal (&)
Institute of Entomology, Academy of Sciences, and Faculty of
Biological Sciences, University of South Bohemia,
BranisÏ ovska 31, 370 05 CÏeske BudeÏjovice, Czech Republic
Fax: +420-38/43625; e-mail: [email protected]
K. Scheller
Theodor-Boveri-Institute, University WuÈrzburg, Am Hubland,
8700 WuÈrzburg, Germany
silk thread (Sehnal and Akai 1990). In the waxmoth
Galleria mellonella (superfamily Pyraloidea) and the
domestic silkworm Bombyx mori (Bombycoidea), the
silk ®bre core, which is secreted in the posterior section
of the silk glands, consists of heavy-chain ®broin and
three other proteins that range in size from 25 to 30 kDa
(Grzelak et al. 1988). In the SDS-PAGE pro®les of silk
extracts treated with 2-mercaptoethanol, the small
components appear as 25-, 27- and 30-kDa bands in
Bombyx and as 25-, 29- and 30-kDa bands in Galleria.
The 25-kDa component of Bombyx was named lightchain ®broin; in native silk it is linked by disul®de
bridges to the heavy-chain ®broin and this complex
molecule was originally regarded as a single protein, the
®broin. Homology of the Galleria 25-kDa silk protein
with Bombyx light-chain ®broin was established by
comparing amino acid sequences of the two proteins
(ZÏurovec et al. 1995).
Couble et al. (1983), however, described a cDNA that
hybridized speci®cally with a 1.1-kb mRNA from the
posterior silk gland section of Bombyx, and showed that
it was translated into a 25-kDa peptide. The authors
named the product silk protein P25 and assumed that it
was identical with the 25-kDa silk fraction. Genomic
clones of P25 were then isolated (Couble et al. 1985) and
sequencing con®rmed that the putative translation
product consists of 220 amino acids (Chevillard et al.
1986a). Chevillard et al. (1986b) recognized, however,
that the deduced amino acid composition of this product
did not match that established by direct amino acid
analysis of the 25 kDa silk fraction. It was proposed that
this fraction contained more proteins, possibly P25 and
light-chain ®broin. This proved not to be the case: immunological studies revealed that the 25-kDa silk component was the product of light-chain ®broin gene,
whereas the P25 gene yielded the 27- and 30-kDa silk
proteins (Tanaka et al. 1993). A similar relationship was
disclosed by partial amino acid sequencing of the puri®ed silk proteins in Galleria ± the silk fraction of 25 kDa
was identi®ed as light-chain ®broin, whereas the 29- and
30-kDa proteins were both found to contain an N-ter-
265
minal sequence similar to that of Bombyx P25 peptides
(ZÏurovec et al. 1995).
The occurrence of two P25 proteins in the gel pro®les of the silk extracts contrasted with the identi®cation of a single P25 mRNA in Bombyx (Couble et al.
1983). In addition, both proteins were considerably
larger than the putative translation product of the P25
gene (Chevillard et al. 1986a). In search of an explanation for these discrepancies we have isolated a P25
cDNA from Galleria and analyzed the corresponding
mRNA and proteins. The results reported below show
that a single translation product is converted into two
proteins via glycosylation. A comparison of our data
with those for Bombyx strongly indicates that a similar
mechanism operates also in this species. The distribution of conserved amino acids in the P25 peptides of
Galleria and Bombyx suggests that certain regions, including the positions of 8 out of 9 cysteine residues, are
essential for the function of the mature P25 protein in
the silk.
Materials and methods
Tissue preparation
Our present stock of the waxmoth Galleria mellonella L. (Lepidoptera, Pyralidae) was obtained from the Department of Animal
Physiology of the Masaryk University (Brno, Czech Republic),
where it had been grown for about 10 years as an o€shoot of our
original laboratory stock. Waxmoth cultures were maintained at
30° C and the larvae were fed on a semi-arti®cial diet (Sehnal 1966).
Posterior parts of both silk glands were dissected from the wateranaesthetized penultimate and last-instar larvae, whose age was
measured from the preceding ecdysis. In the case of fully grown
last-instar larvae, the middle part of silk glands and the carcass
remaining after the silk gland dissection were also analyzed.
Northern analysis was further performed with RNA from the
adults lacking the silk glands, which degenerate shortly after
pupation. All dissected tissues were rapidly frozen in liquid nitrogen or on dry ice and stored at )80° C.
RNA and genomic DNA analyses
The largest cDNA insert was radiolabelled with [a-32P]dATP using
the Random Prime Labeling kit (Amersham) and employed as a
probe in nucleic acid analyses. For the Northern analysis, total
RNA was fractionated by electrophoresis and blotted onto
Hybond-N membranes (Amersham) as described previously (ZÏurovec et al. 1992). Hybridization was carried out under high-stringency conditions at 65° C in 6 ´ SSC, 0.5% SDS; washing was
done at the same temperature, twice in 2 ´ SSC, 0.2% SDS and
twice in 0.2 ´ SSC, 0.1% SDS. In one series of experiments,
Northern blots of the RNA from posterior silk gland section were
hybridized separately with the 5¢ (328 bp) and 3¢ (322 bp) fragments
of the cDNA.
The content of P25 mRNA was established using a dot-blot
assay. Total RNA was extracted from the silk glands with the RNA
Clean system (Angewandte Gentechnologie Systeme). Di€erent
amounts of RNA (2 lg, 0.5 lg, and 0.125 lg) were applied with a
BioDot applicator (Bio-Rad) onto Hybond-N membrane that was
subjected to the hybridization procedure described above for the
Northern analysis. Relative amounts of hybridized RNA were assessed by scanning the developed X-ray ®lm with a CAMAG TLC
Scanner II. The maximal content found in slowly mobile prepupae,
i.e. larvae that are ®nishing cocoon spinning, was taken as 100%,
and values established at other times were expressed in per cent
relative to this number.
Radiolabelled cDNA was also used to probe Southern blots of
the genomic DNA according to the protocol speci®ed by ZÏurovec
et al. (1992).
In vitro transcription/translation
P25 cDNA in pBluescript SK()) was transcribed and translated
in vitro using the TNT Coupled Reticulocyte Lysate system
(Promega) with T3 RNA polymerase. A sample of 0.5 lg of cDNA
was added to 25 ll of reaction mixture containing 8 lCi of
[35S]cysteine as label, and the mixture was incubated at 30° C for 90
min. Transcription/translation of Galleria light-chain ®broin cDNA
was run in parallel for comparison. The [35S]cysteine was chosen as
radioactive marker because P25 contains 9 (Fig. 1), and light-chain
®broin 3 (ZÏurovec et al. 1995) cysteine residues. The translation
products were subjected to SDS-PAGE and visualized by ¯uorography. The molecular weights of the products were assessed using
standard molecular markers (Serva) and pre-stained markers
(Sigma).
Protein analyses
Isolation of cDNA clones and sequencing
A cDNA library in lambda gt10 was prepared from poly(A)+RNA isolated from posterior silk glands of day 3 last-instar larvae
with the aid of a kit (Amersham). The library was screened with the
degenerate oligoprobe 5¢-GGI CCI GC(GATC) AA(CT) GTI GTI
(AC)GI CC(GATC) CC(GATC) (AC)G(GATC) TT(GATC) GAT
GAT GA-3¢, which was derived from the N-terminal sequence
GPANNVVRPPRLDDD of Galleria 29- and 30-kDa silk ®bre
proteins (ZÏurovec et al. 1995). The probe (10 pmol) was end-labelled with 50 lCi of [c-32P]ATP; free nucleotides were removed by
chromatography on a Sephadex G-25 column. Hybridization was
performed at 45° C in 5 ´ SSPE, 5 ´ Denhardt's solution, 0.5%
SDS, and washing at room temperature twice in 2 ´ SSPE, 0.1%
SDS and twice in 1 ´ SSPE, 0.1% SDS. About 20 000 plaques were
screened and three positive phages were identi®ed. Their inserts
were subcloned in pBluescript SK(+/)) (Stratagene) and sequenced with Sequenase Version 2.0 (Amersham). Radiolabelled
[c-32P]ATP, [a-32P]dATP and [a-35S]dATP were purchased from
Amersham. The sequence was analyzed with the program DNA
Star and with the aid of EMBL network services.
Silk proteins were solubilized by soaking chopped cocoons for 48 h
in a solution containing 10 mM TRIS-HCl (pH 7.0), 2% SDS, 8 M
urea and 5% 2-mercaptoethanol. Insoluble heavy-chain ®broin was
removed by centrifugation (5 min at 60 ´ g) and the dissolved
proteins were analyzed by SDS-PAGE. Proteins electroblotted
onto a nitrocellulose membrane were probed for the presence of
sugars with the lectins concanavalin A, lentil agglutinin, and peanut agglutinin (all from Lectinola, Prague), which were conjugated
to horseradish peroxidase (Grubho€er et al. 1990). Conjugates
bound to the blots were revealed with 3,3-diaminobenzidine.
A portion of the dissolved cocoon proteins was subjected to
chemical deglycosylation with the GlycoFree Chemical kit (Oxford
Glyco Systems); the protein solution was dialyzed against 0.1%
tri¯uoroacetic acid, then thoroughly lyophilized and treated with
tri¯uoromethansulphonic acid, which cleaves both the N- and Olinked oligosaccharides. Deglycosylated proteins were precipitated
with 0.5% ammonium bicarbonate and separated from dissolved
reagents by centrifugation. Samples of the original extract of soluble proteins and of the deglycosylated aliquot were dissolved in
the solubilizing bu€er and subjected to electrophoresis in 20%
polyacrylamide gels.
266
Results
Isolation of P25 cDNA
Three cDNA clones ranging in size from 0.5 to 1 kb
were detected with the oligonucleotide probe, which was
designed based on the N-terminal sequences of the
29- and 30-kDa silk proteins. The sequencing of these
cDNAs revealed that all were derived from the 5¢ end of
the same mRNA species and that none of them included
a poly(A) tail. Further screening of the cDNA library
with one of these clones led to isolation of three additional clones that were also truncated in the 3¢ region.
We surmised that the 3¢ region of the target mRNA
contained multiple oligo(A) sequences and that reverse
transcription during the cDNA preparation was initiated from those preceding the poly(A) tail. Hence, a fulllength cDNA had probably not been produced and
further screening of the library was point less. Most of
Fig. 1 Nucleotide sequence of Galleria P25 cDNA (numbered from
the most distal 5¢ nucleotide of the identi®ed cDNA) and sequence of
the deduced peptide, which is aligned with that of Bombyx P25 (amino
acid residues homologous to Galleria P25 are underlined). Cysteine
residues are printed in italics and the putative N-glycosylation sites are
double underlined. The arrow indicates the signal peptide cleavage site
and the beginning of N-terminal sequences of the mature 29- and
30-kDa silk proteins
the cDNA clones we isolated contained the entire coding
region and this was sucient for our study.
In the longest cDNA clone, a translation initiation
codon was identi®ed in position 18 from the 5¢ end and
the coding sequence continued to the termination codon
TAA at position 672 (Fig. 1). The non-coding tail
comprised 370 bp but no polyadenylation signal was
included. Alignment of the coding sequence with the
homologous part of the Bombyx gene P25 (Chevillard et
al. 1986a) disclosed 55.3% nucleotide sequence identity,
con®rming that the identi®ed cDNA corresponded to
the Galleria P25 gene.
The putative translation product
The open reading frame of the Galleria P25 cDNA encodes a peptide of 119 amino acids. The amino acid
sequence GPANNVVRPPRLDDD, which was established by analysing the N-termini of Galleria 29- and
30-kDa silk proteins (ZÏurovec et al. 1995), was found to
begin in amino acid position 17 of the putative product.
This proved that we had isolated cDNA encoding the
29- and 30-kDa silk proteins, and indicated that signal
peptide cleavage occurred between alanine (position 16)
and glycine (position 17) residues of the primary translation product. The N-terminal sequence of the putative
mature peptide di€ered from the stretch of 15 amino
acids identi®ed by peptide sequencing, in having a cys-
267
teine in place of proline in the second position, and a
leucine in place of aspartate in the last position. This
di€erence was not surprising; we expected ambiguities in
the N-terminal sequencing which was performed with
small amounts of insuciently puri®ed proteins.
Utilization of the reading frame opened by the Met
codon at nucleotide position 18, and the homology of
isolated cDNAs with Bombyx P25 gene, were further
con®rmed by aligning the deduced peptide sequence
with that of the primary translation product of the
Bombyx P25 gene (Fig. 1). We found that Galleria and
Bombyx P25s are 73% similar, and 51% identical in
amino acid sequence. All eight cysteine residues and
several stretches of amino acids of the mature peptides
are conserved. The alignment showed that Galleria P25
di€ers from that of Bombyx by two internal deletions
(one of them in the signal region) and one C-terminal
deletions, and by one internal amino acid insertion.
shown). The blot analysis thus proved that all isolated
cDNAs were truncated reverse transcripts of the same
mRNA. This P25 mRNA was detected in the posterior
silk gland of all developmental stages examined (Fig. 2,
lanes 1±4).
Since the Northern data indicated that the amount of
P25 mRNA in the silk glands ¯uctuated, the relative
content of this mRNA in total RNA was quanti®ed by
dot blot assays in course of the penultimate and ®nal
larval instars. We found that the P25 mRNA content
was low at ecdysis to the penultimate larval instar,
peaked in the feeding larvae of this instar, and decreased
again at the molt to the last larval instar (Fig. 3). It rose
rapidly after ecdysis when the larvae initiated feeding,
and reached a maximum in slowly mobile prepupae at
the end of cocoon spinning. The following tenfold drop
of P25 mRNA coincided with the start of silk gland
degeneration.
Northern and dot blot analyses
Analysis of the P25 silk proteins
The following experiments were done to establish the
tissue speci®city of the expression of Galleria P25, and to
ascertain that the 29- and 30-kDa silk proteins are not
encoded by di€erent P25 mRNAs. Northern analysis of
total RNA from the selected tissues demonstrated that
the P25 cDNA hybridized exclusively with a single 2-kb
mRNA from the posterior section of silk glands (Fig. 2,
lanes P, M, C, I). Identical results were obtained when
the total RNA from posterior silk gland was probed
with the 5¢ and 3¢ fragments of the cDNA (data not
The presence of considerable amounts of P25 mRNA in
the silk glands of spinning larvae indicated that representation of the corresponding proteins in the silk would
be high. Indeed, SDS-PAGE analysis of the solubilized
cocoons con®rmed that the P25 proteins of 29- and
30-kDa belonged to the dominant silk components
(Fig. 4, lane 2). In contrast to the previously analyzed
Galleria silk, which contained a single protein in the
30 kDa region (KodrõÂ k 1992), the larvae of our present
stock produced two proteins whose size was close to
30 kDa. The remaining major bands seen in lane 2 of
Fig. 4 represented the 25-kDa light chain ®broin (ZÏurovec et al. 1995) and a 17.5-kDa protein called seroin
(ZÏurovec et al. 1996). Heavy-chain ®broin was not included in the present analysis.
Fig. 2 Northern analysis of P25 expression. Demonstration of P25
mRNA in the posterior silk gland extract (lane P); no signal is found
in identical amounts of total RNA extracted from the middle silk
gland (lane M), larval body carcass (lane C) or whole adults (lane I).
Detection of P25 mRNA in RNA extracts from the posterior silk
glands of larvae dissected in mid-penultimate instar (lane 1) and on
days 1 (resumption of feeding), 5 (termination of feeding) and 6.5
(cocoon spinning) of the last instar (lanes 2, 3, and 4, respectively).
Positions of molecular weight markers are shown on the right
Fig. 3 Changes in the relative amounts of P25 mRNA in the course
of the penultimate and last larval instars as established by dot blot
assay (average values S.D. of 4±5 measurements)
268
Western blots of solubilized cocoon proteins. Several
types of glycosylation were identi®ed in the sericin
fractions above 45 kDa (Fig. 5.) The 30-kDa protein
reacted strongly with concanavalin A (Fig. 5, lane 1),
whereas the 29-kDa protein was not clearly recognized
by any of lectins tested. Chemical deglycosylation
caused the loss of all lectin reactions (Fig. 5, lanes 2, 4,
and 6). Similar results were obtained when the 29- to
30-kDa fraction was electroeluted from the gel, treated
with bacterial glycosidase F, which speci®cally cleaves
N-glycosides, and analyzed by PAGE (data not shown).
Lack of glycosylation in light chain ®broin was con®rmed in the similarly eluted 25 kDa fraction of the
cocoon extract (unpublished data).
Discussion
Fig. 4 Electrophoretic analysis of P25 and light-chain ®broin. SDSPAGE of peptide size markers (lane 1); cocoon extract with major
bands at 30, 29, 25, and 17.5 kDa (lane 2); cocoon extract after
chemical deglycosylation; note the absence of the 30-kDa band, the
faint 29-kDa band, and the strong 25-kDa band (lane 3); products of
in vitro translation of P25 cDNA (lane 4), and light-chain ®broin
cDNA (lane 5). Proteins in lanes 1±3 were stained with Coomassie
Briliant Blue, peptides synthesized in vitro were radiolabelled with
[35S]cysteine and visualized by ¯uorography
While the silk contained P25 proteins of 29 and
30 kDa, the deduced molecular weight of the secreted
translation product (lacking the signal peptide) of the
Galleria P25 mRNA was only 23 098 Da (Fig. 1). In an
attempt to clarify this discrepancy, we decided to analyze the in vitro translation product of P25 cDNA and
to examine if the P25 proteins present in the cocoon
were post-translationally modi®ed. We found that P25
cDNA generated in vitro a single peptide of about
25 kDa (Fig. 4, lane 4), consistent with the estimated
size of 24 863 Da. Our assessment of the size of the
in vitro product was corroborated by a comparison with
the product of the light-chain ®broin cDNA (Fig. 4, lane
5), which encodes a peptide of 25 442 Da (ZÏurovec et al.
1995). As expected, the mobilities of peptides obtained
with P25 cDNA and light-chain ®broin cDNA were
similar (Fig. 5).
Since the sequence of P25 contained several potential
glycosylation sites (see Discussion), the size increase of
the secreted products could be caused by attached sugar
moieties. This proved to be the case, for chemical deglycosylation of the crude mixture of solubilized cocoon
proteins led to a complete disappearance of the 30-kDa
bands and to a pronounced reduction in the intensity of
the 29-kDa band (Fig. 4, lane 3). The remaining 25-kDa
band probably contained both light-chain ®broin and the
deglycosylated 29- and 30-kDa proteins; the relatively
low staining intensity of the 25-kDa band in lane 3 was
obviously due to considerable protein loss associated
with the deglycosylation procedure.
The nature of the glycosylation adducts in the 29- and
30-kDa proteins was investigated using lectin binding to
Using an oligonucleotide derived from a partial sequence of two G. mellonella silk proteins, we identi®ed in
a silk gland-speci®c cDNA library a homolog of the P25
gene known from B. mori (Fig. 1). This present ®nding
of Galleria P25 and our earlier description of the Galleria light-chain ®broin gene (ZÏurovec et al. 1995) contrast
with the report of Tamura and Kubota (1988), who
found no small silk proteins associated with heavy-chain
®broin in several saturniid silkmoths. Saturniids belong
to the same superfamily, Bombycoidea, as Bombyx,
whereas Galleria represents a rather distant superfamily
Pyraloidea (Nielsen and Common 1991). From this
limited information it seems that small silk components,
such as P25 proteins, evolved before Pyraloidea and
Fig. 5 Detection of sugar moieties in silk proteins fractionated by
SDS-PAGE. Cocoon extracts were electroblotted onto a nitrocellulose
membrane and treated with lectins conjugated to horseradish
peroxidase. Silk proteins were reacted before (odd lane numbers)
and after chemical deglycosylation (even lane numbers), with
concanavalin A (lanes 1 and 2), lentil agglutinin (lanes 3 and 4), or
peanut agglutinin (lanes 5 and 6). The position of the 30-kDa protein
is indicated on the left, positions of protein markers on the right
269
Bombycoidea separated in phylogeny and were secondarily lost in the saturniids.
The longest of our P25 cDNA clones included a 5¢
non-translated region of 17 nucleotides. From a comparison with Bombyx P25, in which the ATG initiation
codon lies in position +23 (Couble et al. 1985), we assume that the identi®ed non-coding 5¢ end of Galleria
cDNA is nearly complete. The comparison with Bombyx
P25 further proves that the coding region of 671 nucleotides, detected in our cDNAs, encodes the whole P25
peptide. In the 3¢ non-coding region, however, we
identi®ed only a stretch of 370 nucleotides, whereas the
size of Galleria P25 mRNA (2 kb) suggests that the 3¢
tail is about 1.3 kb long. The 3¢ non-coding region of
Galleria P25 mRNA appears to be much longer than the
474-nucleotide 3¢ tail of Bombyx P25 mRNA (Chevillard
et al. 1986a).
Developmental changes in the content of P25 mRNA
in the silk glands of Galleria resemble those reported for
Bombyx (Couble et al. 1983), in that the content declines
at the molt to the last larval instar (Fig. 3). The rise that
occurs after ecdysis is rather gradual in Bombyx,
whereas in Galleria it is steep in the ®rst 24 h and
gradual afterwards, reaching a maximum in larvae that
are about to ®nish cocoon spinning. A similar di€erence
between Bombyx and Galleria was established in the
developmental pro®le of light-chain ®broin mRNA
(Kimura et al. 1985; ZÏurovec et al. 1995). The rapid
post-ecdysial rise of mRNAs encoding silk proteins apparently enables Galleria larvae to produce large volumes of silk nearly continuously (Jindra and Sehnal
1989), while the spinning of Bombyx larvae is largely
restricted to cocoon construction.
Putative translation products of Galleria P25 and
Bombyx P25 cDNAs are composed of 218 and 220
amino acids, respectively. Representations of individual
amino acids are similar, large di€erences being found
only in tryptophan (1.5% in Bombyx, 0% in Galleria)
and histidine (4.4% and 0.9%, respectively). Positions of
about half of all amino acids are conserved, including
those of all 8 cysteine residues in the mature peptide.
Signal peptide cleavage occurs in Galleria between alanine and glycine (positions 16 and 17) in the AlaGlyPro
sequence, which is conserved also in Bombyx. Since
cleavage between alanine and glycine preceding a proline
residue is common (von Heijne 1986), this site is probably used also in Bombyx. Interestingly, the signal peptide sequence exhibits relatively low conservation; only 5
out of 16 (Galleria) or 17 (Bombyx) amino acid residues
are shared by the two species. This contrasts with the
conservation of 12 out of 16 amino acid residues in the
signal peptide of light-chain ®broins (ZÏurovec et al.
1995).
High conservation of a number of amino acid residues in secreted P25 indicates that their positions are
important for the protein function, but no details are
known. For Bombyx P25 it was suggested that cysteines
are not used for covalent linkage to heavy-chain ®broin
(as is the case of light-chain ®broin), but probably allow
P25 to attain a conformation necessary for its interaction with other ®broin components (Tanaka et al. 1993).
The 53.7% identity between Galleria P25 and Bombyx
P25 indicates that P25 proteins are important for silk
formation in the two species. It is interesting to note that
light-chain ®broin, whose signi®cance and role in silk
production were proved in Bombyx (Takei et al. 1987),
exhibits only 40.2% sequence identity between Galleria
and Bombyx.
Sinohara et al. (1971) demonstrated that Bombyx
®broin (heavy-chain was not separated from the lightchain ®broin and the P25 proteins) is associated with
sugars composed of glucosamine and mannose, which
bind to Asn in the SerAsnThr motif. In Galleria, this
motif is present in the heavy-chain ®broin (our unpublished data) but not in the light-chain ®broin (ZÏurovec
et al. 1995) or the P25 peptide (Fig. 1). However, the
latter contains two N-glycosylation motifs of the type
Asn-X-Ser/Thr (X is any amino acid except proline); the
reaction with concanavalin A (Fig. 5) indicated that
mannose-rich N-glycans are actually bound to the asparagine residue in one or both these motifs in case of
the 30-kDa derivative of P25. No sugar moiety was detected with lectin assays in the 29-kDa derivative, but
changes in protein mobility after deglycosylation (lanes 2
and 3 in Fig. 4) disclosed that both the 29- and the
30-kDa silk proteins contain sugars. We conclude that
the primary translation product of Galleria P25 is converted into two silk proteins by di€erential glycosylation,
the nature of which has been elucidated only in part.
The deduced size of Bombyx P25 is about the same as
in Galleria and the corresponding proteins appear in the
SDS-PAGE of silk extracts treated with 2-mercaptoethanol as 27- and 30-kDa bands. Tanaka et al. (1993)
assumed that P25 exhibits unusual electrophoretic behavior but the situation in Galleria indicates that the
occurrence of two ®nal products and the discrepancy
between the deduced and the apparent molecular
weights in Bombyx P25 proteins is due to glycosylation.
Bombyx P25 contains three N-glycosylation sites, one of
them in an identical position to one of the motifs in the
Galleria P25 (Fig. 1).
Acknowledgements We wish to thank Dr. L. Grubho€er, Institute
of Parasitology AV CÏR, for the supply of lectin-horseradish peroxidase conjugates and for advising us on the methods of glycoprotein analysis. We further acknowledge the gift of Galleria
mellonella larvae from Mgr. Jana BenesÏ ova of the Masaryk University, Brno. Our study was supported by grant 204/96/1100 of the
Grant Agency of the Czech Republic and by Volkswagenstiftung
grant I68 514.
References
Chevillard M, Couble P, Prudhomme JC (1986a) Complete nucleotide sequence of the gene encoding the Bombyx mori silk
protein P25 and predicted amino acid sequence of the protein.
Nucleic Acid Res 14:6341±6342
Chevillard M, Deleage G, Couble P (1986b) Amino acid sequence
and putative conformational characteristics of the 25 kD silk
protein of Bombyx mori. Sericologia 26:435±449
270
Couble P, Moine A, Garel A, Prudhomme JC (1983) Developmental variations of a non®broin mRNA of Bombyx mori silk
gland, encoding for a low-molecular-weight silk protein. Dev
Biol 97:398±407
Couble P, Chevillard M, Moine A, Ravel-Chapuis P, Prudhomme
JC (1985) Structural organization of the P25 gene of Bombyx
mori and comparative analysis of its 5¢ ¯anking DNA with that
of the ®broin gene. Nucleic Acid Res 13:1801±1814
Grubho€er L, Guirakhoo F, Heinz FX, Kunz C (1990) Interaction
of tick-borne encephalitis virus protein E with labelled lectins.
In: Kocourek J, Freed DL (eds) Lectins: biology, biochemistry,
clinical biochemistry, vol 7. Sigma Chemical Company, St.
Louis, Mo., pp 313±319
Grzelak K, Couble P, Garel A, Kludkiewicz B, Alrouz H (1988)
Low molecular weight silk proteins in Galleria mellonella. Insect
Biochem 18:223±228
Jindra M, Sehnal F (1989) Larval growth, food consumption, and
utilization of dietery protein and energy in Galleria mellonella.
J Insect Physiol 35:719±724
Kimura K, Oyama F, Ueda H, Mizuno S, Shimura K (!985) Molecular cloning of the ®broin light chain complementary DNA
and its use in the study of the expression of the light chain gene
in the posterior silk gland of Bombyx mori. Experientia
41:1167±1171
Kodrik D (1992) Small protein components of the cocoons in
Galleria mellonella (Lepidoptera, Pyralidae) and Bombyx mori
(Lepidoptera, Bombycidae). Acta Ent Bohemoslov 89:269±273
Nielsen ES, Common IFB (1991) Lepidoptera (moths and butter¯ies). In: Naumann JD (ed) The Insects of Australia (2nd edn)
Melbourne University Press, Carlton (Victoria, Australia)
pp 817±915
Sehnal F (1966) Kritisches Studium der Bionomie und Biometrik
der in verschiedenen Lebensbedingungen gezuÈchteten Wachs-
motte, Galleria mellonella L. (Lepidoptera). Z Wiss Zool
174:53±82
Sehnal F, Akai H (1990) Insect silk glands: their types, development, and function, and e€ects of environmental factors and
morphogenetic hormones on them. Int J Insect Morphol Embryol 19:79±132
Sinohara H, Asan Y, Fukui A (1971) Glycopeptides from silk
®broin of Bombyx mori. Biochem Biophys Acta 237:273±279
Takei F, Kikuchi Y, Kikuchi A, Mizuno S, Shimura K (1987)
Further evidence for importance of the subunit combination of
silk ®broin in its ecient secretion from the posterior silk gland
cells. J Cell Biol 105:175±180
Tamura T, Kubota T (1988) A determination of molecular weight
of ®broin polypeptides in the saturniid silkworms, Antheraea
yamamai, Antheraea pernyi and Philosamia ricini by SDSPAGE. In: Akai H (ed) Wild Silkmoth `88. National Institute of
Sericulture and Entomological Science, Tsukuba, Japan, pp 67±
72
Tanaka K, Mori K, Mizuno S (1993) Immunological identi®cation
of the major disul®de-linked light component of silk ®broin. J
Biochem 114:1±4
von Heijne G (1986) A new method for predicting signal sequence
cleavage sites. Nucleic Acids Res 14:4683±4690
ZÏurovec M, Sehnal F, Scheller K, Kumaran AK (1992) Silk gland
speci®c cDNAs from Galleria mellonella L. Insect Biochem Mol
Biol 22:55±67
ZÏurovec M, VasÏ kova M, Kodrõ k D, Sehnal F, Kumaran AK
(1995) Light-chain ®broin of Galleria mellonella. Gen Mol Biol
247:1±6
ZÏurovec M, Yang C, Sehnal F, Scheller K (1996) A new component
of the silk ®bre produced in the middle and posterior silk
glands. Proceedings XX International Congress on Entomology, p 425